what is this, I've been here all year and I often see references to it

[b]QB's [/b] E. MANNING / KITNA / CUTLER
[b]RBs[/b] R BROWN / WESTBROOK (R BUSH / MCGAHEE / L BETTS / D WILLIAMS )
[b]WRs[/b] ROY WILLIAMS / T OWENS / A JOHNSON / BERRIAN
[b]TE[/b] WITTEN / WATSON
[b]K[/b] GRAHAM
[b]DT[/b] PANTHERS

[b]QB's [/b] E. MANNING / KITNA / CUTLER
[b]RBs[/b] R BROWN / WESTBROOK (R BUSH / MCGAHEE / L BETTS / D WILLIAMS )
[b]WRs[/b] ROY WILLIAMS / T OWENS / A JOHNSON / BERRIAN
[b]TE[/b] WITTEN / WATSON
[b]K[/b] GRAHAM
[b]DT[/b] PANTHERS

About 3 season ago I got to thinking that while individual fantasy performance is somewhat unpredictable, the nature of playing team defenses is such that there is a certain degree of mathematical predictability to how certain defenses will perform against certain offenses... Being somewhat of a stat junky and math geek I decided to develop a mathematical formula to predict how fantasy defenses will perform against the offenses they face every week... The formula started out fairly simple (as posted above), but over the course of 3 seasons I have tweaked it dozens of times to account for an ever expanding group of variables...

The theory is simple - why play one defense all year long (good matchups and bad) when you can play a different defense off the waiver wire each week based on their matchup...

The current incarnation of the formula takes into account 10 variables per matchup, and I post the results in a stickied thread each week in the "Who to Start" forum.

I can say, ive been using this formula, and for the most part it works. The key is learning to identify the "trap" games, where more than just the #s are concerned.

For instance, as plindsey points out in this weeks thread, the oakland/KC game could be a trap game.

Oakland is bad, and statistically, the teams they play usually put up good fantasy #s. But, Oakland vs KC is typically always a tuff game, lots of scoring, etc, despite how each team is doing. Its one of those games where you can "throw the records out the door". If you can indentify those games, which mathmatically there really is no way to account for, then the system works great. I started Seattle D this past week based on those rankings, and it got me 19 points.

So, while its not foolproof, its a good reference if you dont have chicago or Baltimore as your Defense.

Now, if only someone could come up with a system to rank place kickers weekley. Plindsey, not sure if you ever thought about it, but are there teams that give up more FGs than others? say a few teams that cant seem to stop anyone, but once they get in the redzone, the defense cracks down and holds the offense to 3?

biju wrote:If you, or anyone else reading this thread, is interested in how he comes up with these numbers here's the thread (look about half-way down the page):

biju wrote:If you, or anyone else reading this thread, is interested in how he comes up with these numbers here's the thread (look about half-way down the page):

biju wrote:If you, or anyone else reading this thread, is interested in how he comes up with these numbers here's the thread (look about half-way down the page):